This notebook plots a number of the metric sky maps for a series of opsim runs in single figures as subplots.¶
All of the two year runs were WFD only and the requested number of visits was not decreased. The 10 year runs were also WFD only.
There are 4 major sections of this notebook:¶
- 10 year runs vs 2 year runs
- Airmass opsim runs (Airmass bonus vaired, hour angle bonus 0)
- Hour Angle opsim runs (Airmass bonus 0, hour angle bonus varied)
- Opsim runs with both-bonuses varied
Within each section you will find the following maps:¶
- Pointing history on Alt Az Sky, all filters
- NVisits on Ra Dec Sky, r band
- Median single visit depth, r band
Summary¶
The goal of producing these maps was to highlight the non-uniformity in sky coverage that occurs in the 2 year runs when the airmass bonus and hour angle bonus are increased.
The non-uniformity is espeically well illustrated in the NVisits sky maps when either bonus value is greater than 0.5.
Comparing a 10 and 2 year run with the same bonus values shows that the non-uniform sky coverage is not present in 10 year.
In [1]:
import pandas as pd
from __future__ import print_function
import os
import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline
from lsst.sims.maf.db import OpsimDatabaseV4
import lsst.sims.maf.db as db
import lsst.sims.maf.metrics as metrics
import lsst.sims.maf.slicers as slicers
import lsst.sims.maf.metricBundles as metricBundles
import lsst.sims.maf.plots as plots
import lsst.sims.maf.utils as utils
from lsst.sims.maf.web import MafRunComparison
In [2]:
def subplotSkyMaps(dataframe,outDirs,metricName,nrows,ncols,slicerAbrv='HEAL',
metadata=None,opsim=False,titleList=None,labelList=None,figsize=None,
colorbarRun=None):
runNames = dataframe.index
bundles = {}
userPlotDict = {}
userPlotDict['figsize'] = figsize
if slicerAbrv == 'HEAL':
p = plots.HealpixSkyMap()
if ((slicerAbrv == 'HEAL') & ('Alt_Az' in metricName)):
userPlotDict['rot'] = (0, 90, 0)
if slicerAbrv == 'OPSI':
p = plots.BaseSkyMap()
for i,run in enumerate(runNames):
if metadata:
metadatastr = '_'+metadata
else:
metadatastr = ''
if opsim:
metricFile = 'opsim'+'_'+metricName+metadatastr+'_'+slicerAbrv+'.npz'
else:
metricFile = run+'_'+metricName+metadatastr+'_'+slicerAbrv+'.npz'
bundles[run] = metricBundles.createEmptyMetricBundle()
bundles[run].read(run+outDirs[i]+metricFile)
fig = plt.figure(1, figsize=userPlotDict['figsize'])
if colorbarRun:
userPlotDict['colorMin'] = bundles[run].metricValues.compressed().min()
userPlotDict['colorMax'] = bundles[run].metricValues.compressed().max()
for k,runPlot in enumerate(runNames):
userPlotDict['sub'] = (nrows,ncols,k+1)
if metadata:
userPlotDict['xlabel'] = bundles[runPlot].metric.name+' ('+metadata+')'
else:
userPlotDict['xlabel'] = bundles[runPlot].metric.name
if titleList:
userPlotDict['title'] = titleList[k]
else:
userPlotDict['title'] = "%s %.2f %.2f" % (runPlot,dataframe.loc[runPlot]['airmass_bonus'],
dataframe.loc[runPlot]['hour_angle_bonus'])
p(bundles[runPlot].metricValues, slicer=bundles[runPlot].slicer, userPlotDict=userPlotDict,fignum=1)
plt.suptitle(bundles[runPlot].displayDict['caption'])
fig.subplots_adjust(hspace=0.2)
In [3]:
runs10 = pd.read_csv('colossus_10.csv')
runnames10 = []
for r in runs10['Colossus run']:
runnames10.append('colossus_' + str(r))
In [4]:
comp10 = MafRunComparison(baseDir='.', runlist=runnames10)
In [5]:
stats_10_df,metricDict10 = comp10.combineSummaryStats('WideFastDeep',['airmass_bonus','hour_angle_bonus'])
10 year runs vs 2 year runs¶
In [6]:
subplotSkyMaps(stats_10_df,['/sched/']*len(stats_10_df.index),'NVisits_Alt_Az',nrows=4,ncols=2,
metadata=None,slicerAbrv='HEAL',opsim=False,figsize=(18,30))
NVisits on Ra Dec Sky, r band¶
In [7]:
subplotSkyMaps(stats_10_df,['/sched/']*len(stats_10_df.index),'NVisits',nrows=4,ncols=2,
metadata='r_band_all_props',slicerAbrv='OPSI',opsim=False,figsize=(18,30))
Median single visit depth, r band¶
In [8]:
subplotSkyMaps(stats_10_df,['/sched/']*len(stats_10_df.index),'Median_fiveSigmaDepth',nrows=4,ncols=2,
metadata='i_band_WFD',slicerAbrv='OPSI',opsim=False,figsize=(18,30))
In [9]:
runs_df = pd.read_csv('colossus_runs.csv')
In [10]:
# Create list of runnames from the run_df dataframe
runnames = []
for r in runs_df['Colossus run']:
runnames.append('colossus_' + str(r))
In [11]:
comp = MafRunComparison(baseDir='.', runlist=runnames)
stats_df,metricDict = comp.combineSummaryStats('WideFastDeep',['airmass_bonus','hour_angle_bonus'])
In [12]:
airmass_df = stats_df[(stats_df.airmass_bonus >= 0) & (stats_df.hour_angle_bonus == 0.0 )]
hourangle_df = stats_df[(stats_df.airmass_bonus == 0) & (stats_df.hour_angle_bonus >= 0.0 )]
Airmass opsim runs¶
In [13]:
subplotSkyMaps(airmass_df,['/sched/']*len(airmass_df.index),'NVisits_Alt_Az',nrows=4,ncols=3,
metadata=None,slicerAbrv='HEAL',opsim=False,figsize=(25,25),colorbarRun='colossus_2203')
plt.savefig('airmass_bonus_altaz.pdf')
In [14]:
subplotSkyMaps(airmass_df,['/sched/']*len(airmass_df.index),'NVisits',nrows=4,ncols=3,
metadata='r_band_all_props',slicerAbrv='OPSI',opsim=False,figsize=(25,25),
colorbarRun='colossus_2203')
plt.savefig('airmass_bonus_Nvisits_rband.pdf')
In [15]:
subplotSkyMaps(airmass_df,['/sched/']*len(airmass_df.index),'Median_fiveSigmaDepth',nrows=4,ncols=3,
metadata='r_band_WFD',slicerAbrv='OPSI',opsim=False,figsize=(25,25),colorbarRun='colossus_2203')
plt.savefig('airmass_bonus_depth_rband.pdf')
Hour angle opsim runs¶
In [16]:
subplotSkyMaps(hourangle_df,['/sched/']*len(hourangle_df.index),'NVisits_Alt_Az',nrows=4,ncols=3,
metadata=None,slicerAbrv='HEAL',opsim=False,figsize=(25,25),colorbarRun='colossus_2203')
plt.savefig('hourangle_bonus_altaz.pdf')
In [17]:
subplotSkyMaps(hourangle_df,['/sched/']*len(hourangle_df.index),'NVisits',nrows=4,ncols=3,
metadata='r_band_WFD',slicerAbrv='OPSI',opsim=False,figsize=(25,25),colorbarRun='colossus_2215')
plt.savefig('hourangle_bonus_Nvisits_rband.pdf')
In [18]:
subplotSkyMaps(hourangle_df,['/sched/']*len(hourangle_df.index),'Median_fiveSigmaDepth',nrows=4,ncols=3,
metadata='r_band_WFD',slicerAbrv='OPSI',opsim=False,figsize=(25,25),colorbarRun='colossus_2215')
plt.savefig('hourangle_bonus_depth_rband.pdf')
Opsim runs with both bonuses varied¶
In [19]:
combo_df = stats_df.loc[['colossus_2216','colossus_2217']]
Pointing history on Alt Az Sky, all filters¶
In [20]:
subplotSkyMaps(combo_df,['/sched/']*len(combo_df.index),'NVisits_Alt_Az',nrows=1,ncols=2,
metadata=None,slicerAbrv='HEAL',opsim=False,figsize=(18,6))
NVisits on Ra Dec Sky, r band¶
In [21]:
subplotSkyMaps(combo_df,['/sched/']*len(combo_df.index),'NVisits',nrows=1,ncols=2,
metadata='r_band_WFD',slicerAbrv='OPSI',
opsim=False,figsize=(18,6))
Median single visit depth, r band¶
In [22]:
subplotSkyMaps(combo_df,['/sched/']*len(combo_df.index),'Median_fiveSigmaDepth',nrows=1,ncols=2,
metadata='r_band_WFD',slicerAbrv='OPSI',opsim=False,figsize=(18,6))
